higher priority, always gives the S enantiomer, whereas reversal of this

Jun 24, 2005 - higher priority, always gives the S enantiomer, whereas reversal of this order gives the R enantiomer. Further studies to evaluate the ...
0 downloads 0 Views 260KB Size
6510 Ph

dation, and the donors of the Petroleum Research Fund, administered by the American Chemical Society, is gratefully acknowledged. We wish to thank Professor Ronald Breslow for helpful discussions. A. I. Meyers,* Gerald Knaus Department of Chemistry, Colorado State University Fort Collins, Colorado 80521 Received June 21, 1974

Ph

4

P-Oxido Carbenoids as Synthetic Intermediates. A Facile Ring Enlargement Reaction 6

5a

11 Ph

R'-X

H+.l

(S)-3(R> R)

...Li-OMe 5b

higher priority, always gives the S enantiomer, whereas reversal of this order gives the R enantiomer. Further studies to evaluate the nature of this process revealed that the asymmetric synthesis was independent of the temperature at which the proton was removed from 4. Thus, lithio salts 5 (a and b) were generated at -22, -45, and -78" but were all alkylated at -78" to ultimately give (5')- or (R)-3 in comparable enantiomeric purity. On the other hand, alkylation of 5 (a and b) at various temperatures gave significantly different optical yields of 3 (Table 11). That lower temperatures increase stereoselectivity is no surprise since the AAG for the competing transition state would be enlarged.* The above results now allow a preliminary suggestion regarding the mechanism of this asymmetric synthesis. The proton removal from 4 gives rise to two isomeric lithiooxazolines 5 (a and b) which are probably interconvertible. Attack by the electrophile must be assumed t o occur from the bottomside since it would be difficult to rationalize topside approach in view of the profound effect of methyl sulfate us. methyl iodide. The rate of alkylation on 5a must be faster than that in 5b due to diminished steric interaction between the incoming R group and the R group already present. For methyl sulfate (or tosylate) a bulkier complex (sa) between the lithium cation and the oxygen atoms over that formed with methyl iodide results in widening the relative rates of attack on 5a and 5b causing greater stereoselectivity. If, as assumed, alkylation consistently occurs from the bottomside, this would lead to 6 which upon acidic cleavage provides the carboxylic acids 3. All facts in hand to date are consistent with this mechanism and although further refinements are still necessary, it provides a useful working hypothesis t o allow rational planning of future s y n t h e s e ~ . ~ Acknowledgment. Financial support from the National Institutes of Health, the National Science Foun-

*

(4) 5. D. Morrison and H. S. Mosher, "Asymmetric Organic Reactions," Prentice-Hall, Englewood Cliffs, N. J., 1971, p 40. (5) All mono- and disubstituted oxazolines are new compounds and have been satisfactorily characterized by spectroscopic and elemental analytical methods.

Sir : A number of important ring enlargement reactions have been devised within the last 10 years for the synthesis of macrocyclic derivatives. These reactions assume an even greater importance with the increasing availability of certain medium- and large-ring compounds. The best current methods for a one-carbon ring enlargement involve (a) diazomethane reaction3 or (b) the Tiffeneau-Demjanov r e a r r a ~ ~ g e m e n t Both .~ methods are of limited value in preparative work because of either the variety of products formed (for a) or the number of steps involved (for b).j We report herein a highly effective method for circumventing such difficulties based on the use of P-oxido carbenoids.6 Treatment of the dibromide 1 in tetrahydrofuran with 2 equiv of n-butyllithium in hexane at - 78 " for 30 min and 0" for 5 min led to 89 % isolated yield of pure cyclotridecanone (5). Evidently the P-oxido carbenoid 2 was first formed and decomposed smoothly to the enolate 4, probably via 0-oxido carbene 3' (Scheme I). Similarly, cyclohexanone, cycloheptanone, cyclooctanone, and cyclononanone were prepared from the cor70, 8,10 80," and 87 responding dibromides in 92,8$9 yields, respectively. Clearly, the principal side reaction in this case would be due to replacement of hydrogen by lithium" followed by decomposition to P-oxido monobromocarbene which would produce bromocyclotridecanone after hydrolysis.6 It soon became apparent, however, that the major reaction pathway in the present case was (1) For reviews see (a) C. D. Gutsche and D. Redmore, "Carbocyclic Ring Expansion Reactions," Academic Press, New York, N. Y., 1968; (b) P. R. Story and P. Busch, Advan. Org. Chem., 8,67 (1972). (2) G. Wilke, Pure Appl. Chem., 17,179 (1968). (3) (a) C. D. Gutsche, Org. React., 8, 364 (1954); (b) ref la, Chapter 4. (4) P. A. S. Smith and D. R. Baer, Org. Reacf., 11, 154 (1967); see, also, D. A. Evans, G. L. Carroll, and L. I